Tips for writing an Internet news poll

Today – August 8, 2018 – there’s a poll on the News24 website on the topic of breastfeeding. It asks:

Is breastfeeding in public scandalous?

  Yes, it’s something that should be done in private

  No, but cover yourself

  No, it’s natural and breasts aren’t just for men’s sexual pleasure

The results of a News24 poll on breastfeeding

The results of the News24 poll.

The poll is just one way of asking a question about a burning issue – using what I think of as the “putting words in people’s mouths” option.

On the same day, MSN South Africa is using the same option for the poll choices:

Could SA become the next Zimbabwe?

Yes, we are well on our way

No, people are afraid of change

It’s already too late for us

Things will never ‘get that bad’

I’m ready to emigrate

The results of an MSN South Africa poll on South African politics

The results of the MSN South Africa poll

W24 has the other kind of poll, the one where the vote options are narrowed down simple yes/no choices – it is asking:

Are you attending the #TotalShutdown march?

Yes

No

The results of a Woman24 poll on the Total Shutdown march.

The results of the Woman24 poll.

All three are framed in a multiple choice format, and you don’t see the results of the poll until you have voted.

None of these polls will yield a scientific result, of course – they are really just little bits of Internet fluff, designed to draw readers’ attention to an issue, or drive them to read a story, or perhaps just to generate a few more clicks.

Yet they are all over the Internet, and if they are part of the furniture, then why not do them well?

I have a seat-of-the-pants set of guidelines to making polls. Here they are*:

  1. The poll should be topical, based on the site’s understanding of the news of the day – and it should be changed as soon as it becomes outdated (the W24 poll above refers to an event that has already happened… not such a good idea).
  2. The options should be clear and easy for readers to understand. And broadly applicable. I’d argue that the third option in the News24 poll doesn’t reflect a broad consensus. Lots of people probably think breast-feeding is natural – but it’s unclear how many of those would venture the opinion “breasts aren’t just for men’s sexual pleasure”. The option is harder to choose when so narrowly framed – so No, it’s natural would probably have done the job.
  3. The options should not overlap – they should be different enough that a genuine choice has to be made. In the MSN poll the options are not clearly different from each other. Yes, we are well on our way, and It’s already too late for us are options on the same continuum. People are afraid of change is hard to understand and people might be ready to emigrate for reasons other than a fear that South Africa is likely to become like Zimbabwe. My options as potential answers might have been a simple Yes/No/Maybe. And if I had to put words in people’s mouths I might have done it like this:

Could SA become the next Zimbabwe?

It’s a distinct possibility

No, we are a great country

We are already the next Zimbabwe!

 

  1. To get nuance right, run your poll past someone else – if they look blank, or bored, then you need to go back to the drawing board.
  2. Above all, put yourself in the reader’s shoes: run the question and options through in your mind as if you were seeing them for the first time. Do they make sense? Do they represent the sort of choices that most people might pick?

 

* I am doing some critique of the three polls I used as examples – not to show the particular sites up, but rather in a spirit of constructive criticism. I have made many flawed polls in my time!

Comments are closed.